Towards an improved earthquake catalog for Italy: the seismicity of the Central Italy area as a case study.

Aiming at producing a refined and homogeneous catalog of the Italian seismicity (starting from 1981), we tested selected relocation procedures as applied to seismicity that occured in the Central Italy area for the 2009-2017 period. The starting catalogue is composed by the events detected and located by analysts on duty in the monitoring room of the National Institute of Geophysics and Volcanology (INGV).

We performed tests with two starting datasets with different P- and S-waves arrival times readings accuracy. The first one, related to the 2016-17 period containing the Central Italy seismic sequence, where we directly used the lower resolution, real-time, travel time dataset generated by the analysts on duty 24h per day, including only INGV stations. The second, covering the same area before the 2009 L'Aquila seismic sequence, is composed of the INGV Seismic Bulletin travel time catalogue with off-line revised P- and S-wave readings and merged with available observations from other regional seismic networks.

For both the catalogues, we first used the NonLinLoc code which computes hypocenter locations with probability density functions and exhaustive information about the location uncertainties for each event. This is important when working with such heterogeneous datasets as ours. In addition, we consider lateral heterogeneities by using a regional multi-gridded 3D P- and S- waves velocity model, including station corrections, in the grid search location process. We used the same approach to produce a refined catalog of the Italian seismicity (CLASS, Catalogue of Absolute Locations) which includes 300,786 hypocenters.

After determining the optimal absolute locations for both catalogues, we re-located all of them with the double-difference method, using cross-correlation phase delay time measurements for improving the resolution in pick based delay times. We are using a subset (2016-2017) of the catalog to test the performance of the relocation process, and for potential implementation of the real-time double-difference system DDRT.